We can think of research as having two phases:

  1. a “creative phase,” during which researchers search for new ideas, and
  2. a “working phase,” during which researchers exert effort on an idea.

Sadler (2021) views these phases as complementary productive inputs, and analyzes how and why the optimal input mix changes when features of the research environment change. For example, if creativity becomes more expensive then researchers spend more time working on each idea and, in Sadler’s model, tend to work on lower quality ideas. Policymakers can use subsidies and taxes to change the relative costs of creativity and work, thereby influencing how researchers allocate their time.

The core features of Sadler’s model are as follows. A researcher wants to maximize the (sum of discounted) payoffs from their life’s work on research ideas. Different ideas have different qualities, and the researcher knows an idea’s quality when they find it. What they don’t know is whether each idea is feasible: some ideas are “dead ends” in the sense that they cannot generate payoffs, no matter their quality or how long the researcher works on them. The longer the researcher works on an idea without it paying off, the more they start to believe the idea is a dead end. The researcher can act on this belief at any time by abandoning an idea and searching for a new one.

In Sadler’s model, the researcher continues working on an idea if and only if the expected payoff exceeds the sum of two costs: (i) the effort required to work on the idea and (ii) the opportunity cost of not searching for another idea. This opportunity cost depends on the researcher’s discount rate for future payoffs because new ideas take time to find. The expected payoff from working on an idea falls over time, implying that there is a unique amount of time that the researcher spends on each idea before abandoning it. This amount of time is larger for ideas with higher quality.

If effort becomes more costly then the researcher spends less time working on each idea and focuses their effort on higher quality ideas. Intuitively, this rise in effort costs makes the creative phase relatively cheaper, so the researcher substitutes towards it. On the other hand, if the researcher’s discount rate rises (i.e., they become less patient) then they spend more time in the working phase and allow themselves to work on lower quality ideas. This is because the discounted payoff from continuing to work on an idea falls by less than the discounted opportunity cost of searching for a new idea.

Sadler’s model helps explain why organizations with a shorter-term focus (i.e., a higher discount rate) tend to be less innovative: they spend too much time working on low quality ideas, and not enough time searching for high quality ideas. In contrast, organizations that use longer-term incentives, such as stock options and golden parachutes, tend to be more innovative (Lerner and Wulf, 2007; Francis et al., 2011).

Sadler’s model also demonstrates how subsidies and taxes can influence how researchers allocate their time. Subsidizing effort during the working phase lowers the relative cost of that phase, and so researchers spend more time working but tend to work on lower quality ideas. Taxing payoffs raises the quality threshold for abandoning an idea, and so researchers spend less time in the working phase but tend to work on higher quality ideas. The optimal policy depends on the social value of research: the more convex is that value in idea quality, the more society wants researchers to focus on fewer but higher quality ideas, and so the more attractive are taxes relative to subsidies.